One day you find yourself feeding the code of a beloved hobby project through a coding agent and getting back a troubling diagnosis: you have a few security holes.

That’s the catch, isn’t it? 

I’ve been writing code since I was in grade school. I’ve made games, apps, sims, and platforms. It transcends dabbling, but outside of professional gigs my hobby code is still often little more than a series of beloved weekend projects.

When you build from scratch—and this is a well-trod warning—two things happen: first, you learn more, but second, you find yourself re-inventing the wheel.  You need to write code to build things that someone already puzzled out years ago. You get smarter, you flex those brain cells, but you are solving a problem that has already been solved… and you’re maybe not solving as well as someone else already did.

Thus: you tend to have a few security holes—as the agent bluntly informed me.

I don’t want to use AI to write my code, really, but in a world where agents are digging through my website—that one, the one I built from scratch and worked through a hundred little coding problems just to get it to finally work—I find myself in desperate need of a strategic alliance. I can’t ignore the fact that to participate online as a developer in this modern era means leveraging systems that can think as quickly and complexly as the systems trying to break through my security.

It’s like the old saying: bringing a knife to a gun fight. 

Bringing a single human brain to an AI fight might be just as foolish.